JPH04241074A - Automatic document clean copying device - Google Patents

Abstract

PURPOSE:To automatically make a clean copy of a handwritten character document without inputting any clean copying data by an operator. CONSTITUTION:Handwritten characters are read by a read part 1 and converted by an A/D conversion part 2 into a digital signal, which is stored in a memory 3. The handwritten character string is cut, line by line, by a character line retrieval part 4, respective characters of each line of the handwritten character string are segmented by a character area segmentation part 5, and segment elements of each character are detected by a character constituent element detection part 6 and stored in a data storage part 7. Then a size varying part 8 varies the longitudinal and lateral lengths of each character to a proper size and a position correction part 9 corrects the lateral positions of respective characters so that the character string have constant lateral intervals; and character strings of respective lines are reconstituted by a document image shaping and reconstitution part 10 so as to have specific intervals, and they are stored in a memory 11 and then outputted through an output part 12.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic document formatting device for automatically formatting documents.

0003

2. Description of the Related Art Conventionally, this type of document copying apparatus has been designed to adjust the vertical and horizontal sizes of image data display areas on documents and image It is configured to perform calculations so that no missing portion occurs in the display area and the aspect ratio of the image data is 1:1 depending on the vertical and horizontal sizes of the data. Therefore, when the operator inputs formatting data such as the vertical size and horizontal size of the image data display area, this document formatting device ensures that the image data does not have any missing parts and that the aspect ratio of the image data is 1:1. You can correct a string by calculating it so that it becomes 1.

0004

However, in the above-mentioned conventional document formatting device, since calculations are made based on the vertical size and horizontal size of the image data display area, the operator must input these formatting data. Another problem is that it is not possible to automatically clean up documents with handwritten characters.

SUMMARY OF THE INVENTION In view of the above conventional problems, an object of the present invention is to provide an automatic document formatting device that can automatically format documents with handwritten characters without the operator inputting any formatting data. do.

[0006]

[Means for Solving the Problems] In order to achieve the above object, a first means includes a reading means for optically reading a handwritten character string, and a reading means for searching the handwritten character string read by the reading means. an extracting means for searching line by line; a cutting means for cutting out the handwritten character string of each line extracted by the searching means character by character; and a determining means for determining the size of each handwritten character cut out by the cutting means; scaling means for scaling the vertical and horizontal lengths of each character discriminated by the discriminating means so that each character has an appropriate size; and a scaling means for scaling the length and width of each character discriminated by the discriminating means so that the intervals between the handwritten character strings in each line extracted by the searching means are appropriate. The apparatus is characterized in that it comprises an arrangement means for arranging the characters scaled by the scaling means, and an output means for outputting the character strings of each line arranged by the arrangement means.

[0007] In the second means, the scaling means of the first invention discriminates and vectorizes the line segment elements of each character discriminated by the discrimination means, and scales the vectorized line segment elements. It is characterized by

[0008] In the third means, the determining means of the second invention recognizes each handwritten character cut out by the cutting means by matching with a dictionary pattern, and determines the outline of the font data by referring to the character font of the dictionary pattern. It is characterized by converting lines into vectors.

[0009]

[Operation] The first means uses the above configuration to extract the handwritten character string line by line, separate the handwritten character string in each line character by character, determine the size of each character, and make the appropriate size. The vertical and horizontal lengths are scaled so that the handwritten character strings in each line are spaced appropriately. Therefore, it is possible to automatically format a document with handwritten characters without the operator inputting any formatting data.

[0010] In the second method, the line segment elements of each character are vectorized, and the vectorized line segment elements are scaled, so that the handwriting as intended by the scribe is corrected.

[0011] In the third means, handwritten characters are recognized by matching with a dictionary pattern, and font data of the dictionary pattern is vectorized. Therefore, since the vectorized line segment elements are scaled, handwritten characters can be formatted into a dictionary pattern character font.

0012

Embodiments Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of an automatic document cleansing device according to the present invention, FIG. 2 is an explanatory diagram showing a handwritten document read by the automatic document cleansing device shown in FIG. 1 and its histogram, and FIG. 3 is a diagram showing the characters in FIG. FIG. 4 is an explanatory diagram showing the line image extracted by the line search unit. FIG. 4 is an explanatory diagram showing the characters cut out by the character area clipping unit in FIG. 1. FIG. 5 is an explanatory diagram showing the constituent elements of the characters in FIG. 4. 6 is a flowchart for explaining the operation of the character component detection section in FIG. 1, FIG. 7 is a flowchart for explaining the operation of the character area size changing section and position correction section in FIG. 1, and FIG. An explanatory diagram showing the scaling operation of the area resizing unit. FIG. 9 is an explanatory diagram showing vector data before and after being scaled by the character area resizing unit in FIG. 1. FIG. 10 is an explanatory diagram showing the document image formatting and reconstruction in FIG. 1. It is an explanatory view showing operation of a part.

In FIG. 1, an image of a handwritten manuscript or the like is optically read by a reading section 1 and converted into an electrical signal, and this electrical signal is converted into a digital signal by an A/D converting section 2 and stored in a memory 3. be done. The left side of FIG. 2 shows an example of a handwritten manuscript P, and this handwritten manuscript P has horizontally written handwritten characters 13 and a rectangular image 14 such as a photograph. The right side of FIG. 2 shows a histogram 15 in which the handwritten characters 13 of the handwritten document P and the rectangular image 14 are accumulated in the horizontal direction, and the histogram 15 in the rectangular photographic image 14 is considered to be uniformly added. Therefore, handwritten characters 13
The break with is saved.

The character line search unit 4 shown in FIG. 1 determines a point where the cumulative frequency of the histogram 15 changes drastically in the vertical direction based on the image data stored in the memory 3, considers it as a cutting point of a character line, and searches the line. Extract the image. The upper part of FIG. 3 shows a line image 17 extracted by the character line search unit 4,
This line image 17 has character areas 16 written horizontally. The lower part of FIG. 3 shows a histogram 18 in which the black pixels of each character area 16 are accumulated in the vertical direction.

The character area cutting unit 5 shown in FIG. The position and vertical and horizontal length of the character part are calculated by removing the margins. FIG. 4 shows the character “A” cut out by the character area cutting unit 5, and this character has an x coordinate Xs of the lower left corner of the character area 16, a y coordinate Ys of the upper right corner, and a horizontal length x length. , vertical length y length, etc. FIG. 5 shows the character components of this character “a”, each character component having a plurality of pixels 25
and the end point 23 of the stroke and the vector 24
is expressed by

As shown in FIG. 6, the character component detection unit 6 shown in FIG. is sampled and the average line width is calculated (step 20). Next, the character component detection unit 6
Detect the end point 23 of the character stroke (step 21)
, the intersection of this end point 23 and the stroke creates a vector 24
The character image is vectorized by calculating (step 22), and the position, size, and vector data of the character area 16 are stored in the data storage section 7.

As shown in FIG. 7, the character area resizing section 8 shown in FIG. 1 first classifies the heights of the cut out characters into a plurality of categories (step 26). That is, for example, the characters are classified into categories such as relatively large characters such as headings, normal-sized characters such as sentences, and relatively small characters such as the consonant "tsu". Next, calculate the average height of characters representing each category, compare this average height with the height of each character to determine whether each character belongs to that category, and assign each character to that category. The image is scaled to the average height (step 27). That is, as shown in FIG. 8, for example, the height y length of the cut out character 32 and the average height y av of the category
The scale is changed by e to a character 34 of height y ave . Next, as shown in FIGS. 7 and 8, the size changing unit 8 compares the aspect ratio of the character (step 28), and increases the horizontal length x length at the same ratio as the vertical scaling factor. The magnification is changed to x reform (step 29). Note that this lateral length x reform can be determined by equation (1). x reform=(x length/y
length)*y ave...(1)

0018
As shown in FIG. 9, the start and end points of the vector data of the character "A" are indicated by addresses in the image area 39 before scaling and in the image area 40 after scaling. Therefore, the amount of displacement in the vertical and horizontal directions of the vector is calculated based on the scaling factors in the vertical and horizontal directions of the cutout character area 39 before scaling, and the position address of the vector is calculated according to the position correction of the cutout area 39. By similarly correcting and scaling each stroke of the character, each stroke of the character area 40 after scaling can be obtained.

Next, the character area position correction unit 9 shown in FIG. 1 calculates the average length of horizontal blanks between each character as shown in FIG. Correction is performed using blanks so that the characters are placed in appropriate positions and the number of characters in one line is the same before and after scaling (step 31). Furthermore, this correction is
The blank length that can exist in the row image 17 is calculated by summing up the average horizontal length x reform of each area after scaling and subtracting this total value from the length of the cut row image 17. , by dividing this blank length by the number of characters to calculate the average blank length between each character area.

Next, the document image shaping and reconstruction unit 10 shown in FIG. 1 converts the character string 36 into a character string 36 such that the bottom of the character strings that do not match in the vertical direction matches the line 35 at a predetermined interval, as shown in FIG. Reconfigure. Note that the distance 37 between characters in the horizontal direction is the average blank length calculated by the position correction unit 9. The image data corrected in this manner is temporarily stored in the memory 11, and is outputted as a print or the like by the output unit 12.

Therefore, according to the above embodiment, the character line search unit 4 cuts out the handwritten character string line by line, the character area cutting unit 5 cuts out each character of the handwritten character string in each line, and the character component detection unit 6 cuts out each character of the handwritten character string in each line. The line segment element of each character is detected, the size change unit 8 determines the size of each character, and the vertical and horizontal lengths 38 are scaled to the appropriate size, and the position correction unit 9 converts the character string The reconstruction unit 10 corrects the horizontal position of each character so that the horizontal spacing is constant.
Since the spacing between the character strings in each line is corrected to be constant, it is possible to automatically format a document with handwritten characters without the operator inputting any formatting data.

[0022] In this case, the number of characters included in one line is the same, and the handwritten characters with a conscious size are scaled according to the size, so the layout is the same as the scribe intended. Output. Furthermore, since the line segments of handwritten characters are vectorized, it is possible to prevent characters from becoming blurred, cut, or added with noise, and furthermore, it is possible to preserve the handwriting of the scribe.

FIG. 11 is a flowchart for explaining the operation of the character component detection section 61 in the second embodiment of the automatic document fairing apparatus according to the present invention, and FIG. 12 is an explanatory diagram showing the scaling operation thereof. . In FIG. 11, a character component detection unit 61 matches each character extracted by the character area extraction unit 5 in the first embodiment with a dictionary pattern, performs structural analysis based on character components, and performs meaning of characters in a sentence. It is recognized by analysis etc. (step 41). Next, the outline of the font data is vectorized by referring to the character font of the dictionary pattern (step 42). Therefore, as shown in FIG. 12, when the size changing section 9 enlarges the font data 43 of the dictionary pattern, characters 44 similar to printed characters are output, so that the handwritten character string can be completely transcribed.

[0024]

As explained above, the invention according to claim 1 includes a reading means for optically reading a handwritten character string, and a reading means for searching the handwritten character string read by the reading means, line by line. extracting means for cutting out the handwritten character string of each line extracted by the searching means for each character; determining means for determining the size of each handwritten character cut out by the clipping means; scaling means for changing the length and width of each identified character so that it becomes an appropriate size; Since it is equipped with an arrangement means for arranging each character scaled by the multiplication means and an output means for outputting the character strings of each line arranged by the arrangement means, the operator does not have to input any data for fair copying. Documents with handwritten characters can be automatically formatted.

[0025] The invention according to claim 2 distinguishes the line segment elements of each handwritten character, vectorizes them, and scales the vectorized line segment elements, so that the scribe can write with the handwriting intended by the scribe. can.

[0026] The invention of claim 3 recognizes each handwritten character by matching it with a dictionary pattern, and vectorizes the outline of the font data by referring to the character font of the dictionary pattern. You can format the font.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of an automatic document fairing device according to the present invention.

FIG. 2 is an explanatory diagram showing a handwritten manuscript read by the automatic document fairing device of FIG. 1 and its histogram.

FIG. 3 is an explanatory diagram showing a line image extracted by the character line search unit of FIG. 1;

FIG. 4 is an explanatory diagram showing characters cut out by the character area cutting section of FIG. 1;

FIG. 5 is an explanatory diagram showing the constituent elements of the characters in FIG. 4;

FIG. 6 is a flowchart for explaining the operation of the character component detection section of FIG. 1;

FIG. 7 is a flowchart for explaining the operations of the character area size changing unit and position correcting unit in FIG. 1;

FIG. 8 is an explanatory diagram showing a scaling operation of the character area size changing section of FIG. 1;

FIG. 9 is an explanatory diagram showing vector data before and after being scaled by the character area size changing unit of FIG. 1;

FIG. 10 is an explanatory diagram showing the operation of the document image shaping and reconstruction section of FIG. 1;

FIG. 11 is a flowchart for explaining the operation of a character component detecting section in a second embodiment of the automatic document fairing apparatus according to the present invention.

FIG. 12 is an explanatory diagram showing the magnification changing operation.

[Explanation of symbols]

1 Reading section 4 Character line search section 5 Character area extraction part 6 Character component detection unit 8 Size change section 9 Position correction section 10 Document image formatting and reconstruction unit 12 Output section 61 Character component detection unit

Claims (3)

[Claims] 1. Reading means for optically reading a handwritten character string; extraction means for searching the handwritten character string read by the reading means line by line; a cutting means for cutting out a handwritten character string character by character; a determining means for determining the size of each handwritten character cut out by the clipping means; scaling means for scaling the vertical and horizontal lengths; and arranging means for arranging the characters scaled by the scaling means so that the intervals between the handwritten character strings in each line extracted by the searching means are appropriate. and an output means for outputting character strings in each line arranged by the arrangement means. 2. The scaling means discriminates and vectorizes line segment elements of each character discriminated by the discriminating means,
2. The automatic document formatting apparatus according to claim 1, wherein the vectorized line segment element is scaled. 3. The discrimination means recognizes each handwritten character cut out by the cutout means by matching it with a dictionary pattern, and vectorizes the outline of the font data by referring to the character font of the dictionary pattern. The automatic document integrator according to claim 1.